Hindawi Publishing Corporation

Advances in Mechanical Engineering

Volume 2013, Article ID 539857, 8 pages
http://dx.doi.org/10.1155/2013/539857

Research Article
Offshore Desalination Using Wave Energy

Álvaro Serna and Fernando Tadeo

Departamento de Ingenierı́a de Sistemas y Automática, Universidad de Valladolid, Calle Doctor Mergelina s/n, 47011 Valladolid, Spain

Correspondence should be addressed to Álvaro Serna; alvarosercan@hotmail.com

Received 27 March 2013; Revised 30 May 2013; Accepted 23 June 2013

Academic Editor: Jamel Orfi

Copyright © 2013 Á. Serna and F. Tadeo. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.

This paper evaluates the design of an offshore desalination plant currently under preliminary development. The purpose is to test
the feasibility of producing drinkable water using wave energy in out-of-sight installations, as an alternative for those locations
where land use, civil engineering works, and/or environmental impact make a coast-based solution inadequate. After describing
the components, a proposal for sizing them is studied, based on using buoy-measured data at the expected location and their
mathematical models of the different sections of the plant. Finally, by using measured buoy data, the influence of sizing on the
expected performance is studied for a specific location, and one of the designs is developed in detail.

1. Introduction with or without connection to the local electricity grid.

Offshore desalination plants powered by renewable energies
are being proposed as an alternative for a coastal desalination
facility, for those locations where the lack of suitable land
makes a land-based desalination plant inadequate [1, 2].
Different techniques of desalination have been studied in
previous works. Among the most developed are reverse
osmosis and vapor compression desalination [3].
This paper studies a proposal to use an offshore wave
platform as the sole energy source of a reverse osmosis (RO)
plant to produce drinkable water, which is then transported
offshore (through pipes, water tank ships, or bladders) [4].
Some economic studies have been published about
the advantages of autonomous wave-powered desalination
plants. These works have discussed whether the real value of
the system is enhanced due to its flexibility for deployment
and reduced environmental impact [5, 6]. As renewable
energy is the only source of energy, a central problem would
be to balance energy consumption with energy production;
as energy production is variable and the electrical system
is isolated from the grid, the desalination plant has to be
designed with variable production in mind [7, 8]. Most
of the works on wave energy conversion have focused on
electricity production. Any such converter could, in principle,
be coupled to an electrically-driven desalination plant, either
Various concepts have associated wave energy converters and
RO [9].
The proposal presented here is based on dividing the
plant into a few sections, which would be switched on
or shutdown, depending on the available energy [10, 11].
All of this requires a specific control system, which is
discussed later. Wave energy converters are studied as they
provide lower variability in energy production in comparison
with other sources [12]. Thus, power consumption adapts
to power production by connecting or disconnecting sec-
tions of the plant (following a Smart Grid approach for
the microgrid in the plant) and using temporary storage
of electricity for short-time balances and an increase in
autonomy.
The process diagram in Figure 1 presents the main
blocks of our proposal: the wave energy converters (WECs),
batteries, the seawater pumps (SWPs), the seawater tank,
the desalination plant, the fresh water tank, and the water
transport facility (by tankers, bladders, or pipes). After the
introduction, a description of the different components is
given in Section 2. A control system is proposed in Section 3.
Section 4 gives some guidelines for the system sizing method-
ology. In Section 5, a case study is presented. Section 6 depicts
some results and finally, some conclusions are provided in
Section 7.
2 Advances in Mechanical Engineering

Waves

Wave energy
converters
Seawater Batteries

Electrical power
Drinkable
water

Seawater Seawater Desalination Fresh water Tanker

pumps tank plant tank bladders

Figure 1: Process diagram.

2000

Power produced (kW)

Heaving buoys
Support structure 1500

1000

Tension wires 500

8
0 6
16 )
14 4 (m
Ballast
12
10 i ght
Peak p 8 2 t he
eriod 6
4 0 ifi can
(s) n
Figure 2: Scheme of the WEC. Sig

Figure 3: WEC power matrix.

2. Description of Components
2.1. Wave Energy Converters. The electrical power generated
by a wave energy converter (WEC) is determined by its
characteristics and the wave weather in the location [13].
We assume here the use of a multibody floating WEC; the
energy is extracted by the relative motion of different parts
of the structure. These devices are adequate for in deep water
(>40 m) [11].
More precisely, the device used is a floating heave-buoy
array (F-HBA), which is represented in Figure 2. It is a
multibody floating WEC, composed of many heaving buoys
connected to a common reference structure. This structure
is composed of an arrangement of a single support structure
and a series of ballast baskets, connected through tension
wires. The total buoyancy force from the buoys is balanced
by net gravity forces of the bridge and the ballast baskets.
The buoys are connected to the submerged structure via a
hydraulic Power Take-Off (PTO) system, which converts the
mechanical energy of the device into electricity. In the case of
wave activated body WECs, they can be based on hydraulic
components (hydraulic rams and motors) combined with an
electrical generator [14], or they can be fully electric [15, 16].
The second proposal was assumed in this work due to the
special conditions in offshore plants. The microgrid designed
in this paper has shedable loads (RO sections and SWP), so
the consumed power adapts to the varying input power. The
battery controllers ensure the needed grid regulation.
One of the key points in the structural design and energy
extraction capacity of the device is the response to different
periods and wave heights (Figure 3). To evaluate the energy
produced by the WEC, water waves are considered to travel
along the surface of the sea with an approximate sinusoidal
profile, characterized in terms of the time between successive
crests and the size of these crests [17]. Wave height and
period are represented by statistical measurements, the most
common being the significant wave height, 𝐻𝑠 (𝑡) (around
four time the root-mean square of the surface elevation
during a given window), and the wave period 𝑇𝑝 (𝑡). The
devices have a maximum range of operation; the energy
that can be used by a device is limited to a maximum wave
height and a minimum wave period. Multiplying the WEC
power matrix (Figure 3) by the buoy-measured data of the
sea location, the mean absorbed power during a specific time
period of the device can be derived (𝑃𝑤 ).
2.2. Water Production. Different techniques for fresh water
production have been developed. Nowadays, reverse osmosis
and distillation processes are the most used technologies.
The required plant capacity, the annual and daily distribution
of fresh water demand, the product cost, the technology
maturity, and any problem related to the coupling of the
renewable energy and the desalination systems determine
this selection [18]. The plant proposed in this paper is an
offshore floating plant, which makes the implementation of
Advances in Mechanical Engineering
Pressurized
seawater
Fresh
water
Brine
Figure 4: Transport of water through an RO membrane.
distillation processes difficult due to the strict limitations
on space and movements. Thus, RO was selected as the
desalination technique as there is wide experience in this
process in marine installations.
RO is based on the transport of water through a mem-
brane by porosity and/or diffusion (see scheme in Figure 4).
The chemical and physical nature of the membrane deter-
mines its ability to allow for preferential transport of the sol-
vent (water) over the solute (salt ions). To ensure water flow
through the RO membrane, a difference of pressure is needed
between the feed and permeate sides of the membrane, which
must be significantly greater than the osmotic pressure; for
seawater this gives operating pressures around 60 bars [19];
the pressure of the brine is frequently recovered using specific
devices to improve energy efficiency.
2.2.1. Desalination Plant Operation. As has already been
mentioned, WEC energy production changes with time,
which requires the power consumption (and thus the produc-
tion of power) to adapt to the available power. Changing the
working point of the plant (thus, changing the flow/pressure
combination from a specific value) makes it possible to adapt
the RO electrical load to the demand (this is challenging for
RO plants, as they are normally designed to operate at an
optimal working point).
A specific control system, discussed later, would be
responsible for adapting power consumption to demand,
always trying to operate the plant at the best possible overall
efficiency.
It was decided to partition the high pressure part of
the RO plant into three sections (A, B, and C) [4]; each
section operates as a standard RO plant (producing water
with maximum efficiency, at a fixed working point). Sections
can operate on an individual basis or together with one or
more of the different sections. Seawater is taken from the
sea by the three independent pumps before being stored in a
common seawater tank as can be seen in Figure 5 (ratings in
this figure correspond to the case study that will be presented
in Section 5). It is then desalinated and stored in a fresh water
tank in order to transport it offshore.
2.3. Short-Term Energy Storage. The proposed system is
offgrid, so energy production and demand must be balanced.
Short-term energy storage is used to provide energy to the
reverse osmosis plant and the SWPs when not enough power
is provided by the WEC [20].
Several technologies have been proposed for temporary
storage (see, e.g., [21] for a review of technologies). In this
case, we will assume a standard Pb battery rack system,
3
as there is wide experience in marine applications. A simple
mathematical model of this system is as follows (the nominal
voltage of the lead-acid cell is 2 volts) [22]:
𝑃𝑊 − 𝑃𝑇
𝐼= , (1)
𝑉
where 𝐼 is the current at the battery terminals, 𝑉 is the voltage,
𝑃𝑊 is the power supplied by the WEC, and 𝑃𝑇 is the total
power consumed by the SWPs and the RO. The equation
relating the capacity of the battery and the amperage is as
follows [11]:
𝑑𝐶𝑅 𝐼𝑘
=𝜂 , (2)
𝑑𝑡 3600
where 𝐶𝑅 is the rated capacity at that discharge rate, 𝑡 is the
time (h), 𝑘 is Peukert’s coefficient, and 𝜂 is the efficiency of
the battery. The key decision variable, the depth of discharge
(DOD) of a battery, is the percentage of capacity to which it
is discharged. Values approaching 1 correspond to the battery
being almost discharged. The DOD is given by the expression
below, where 𝐶𝑃 is Peukert’s capacity:
𝐶𝑅
DOD = 1 − . (3)
𝐶𝑃
The value of the % of battery charge is given by this simple
equation:
% Battery charge = (1 − DOD) ⋅ 100. (4)
It is important to remark the fact that the constant
discharging and charging of the batteries lead them to becom-
ing quickly ineffective, so DOD is an important parameter
that will be used inside the control system, as in [23, 24].
Therefore, the battery capacity that has finally been chosen
in the simulation in Section 6 aims to be the least harmful to
the system.
3. Control System
A central component to ensure the smooth operation of the
facility is the control system, which should balance electricity
production and consumption by adapting the consumed
loads: the power of the seawater pumps (SWPs) and the
pumps working in the different sections of the RO. Battery
controllers regulate the local grid frequency and voltage. The
control system is based on the two following ideas.
Six control variables are used, that correspond to the
connection/disconnection of each of the sections A, B, and
C in the RO and each SWP (see Figure 6).
(i) When the production is higher than the consump-
tion, the energy is stored until the temporary storage
system is nearly full; then different sections of the RO
and the SWPs are switched on.
(ii) When the storage system is almost empty, sections
of the RO and the SWPs are switched off (these sec-
tions are then automatically cleaned and maintenance
operations are carried out).
4 Advances in Mechanical Engineering

C section (RO)
95 kW

SWP 1 Flow (C) = 40 m3 /h

80 m3 /h
45 kW
B section(RO)
150 kW
SWP 2
80 m3 /h Flow (B) = 70 m3 /h
45 kW
A section (RO)
SWP 3 340 kW
80 m3 /h
45 kW Flow (A) = 150 m3 /h

Seawater tank Fresh water tank

Seawater supply Reverse osmosis

Figure 5: Structure of the proposed desalination plant.

Produced Battery
DOD Idle
energy
DOD ≤ 0.8

B DOD ≥ 0.85
Control system
DOD ≤ 0.6

C+B DOD ≥ 0.65

DOD ≤ 0.45
switch on/off SWP 1

switch on/off SWP 2

switch on/off SWP 3

A DOD ≥ 0.5
switch on/off

switch on/off

switch on/off
section A

section C
section B

DOD ≤ 0.3

A+C DOD ≥ 0.35

DOD
≤ 0.2
A+B DOD ≥ 0.25
Reverse
Seawater DOD
osmosis ≤ 0.05
supply
plant A+B
+C
Integer variable
continuous variable DOD ≥ 0.2

Figure 6: Proposed control system. Figure 7: Sequential function chart of the control logic of the
operation.

A control subsystem decides the connection/disconnec- (iv) When the level of the seawater tank is lower than
tion of the three SWPs in specific cases of safety. a certain value, the three SWPs are switched on
simultaneously until the level is again between safety
(iii) When the level of the seawater tank is higher than limits.
a certain value, the three SWPs are switched off
simultaneously until the level is again between safety The control logic is summarized in the Sequential Function
limits. Chart in Figure 7, where the different stages in the operation
Advances in Mechanical Engineering
From C + B
or A + C
A.0
Idle
L ≤ L max and DOD ≤ 0.45
A
SWP 1
L ≤ L max and DOD ≤ 0.4
A
SWP 1
+2
L ≤ L max and DOD ≤ 0.35
A
SWP 1 +
2+3
L ≤ L max and DOD ≤ 0.3
To A + C
Figure 8: Detail of the sequential function chart (stage A).
can be seen. There is one different combination of RO sections
(A, B, and C) in each stage. Therefore, for each range of DOD
values, there is a certain operating stage. Furthermore, in
each stage, the three SWPs are switched on/off one by one,
depending on the value of the DOD, as detailed in Figure 8,
where the Sequential Function Chart in stage A is shown.
4. Sizing
A central aspect of the design is the selection of the sizes of
components. A parallel approach to the one presented in [11]
is presented here. The following parameters are selected.
(i) The fresh water produced, 𝑄𝑇 (m3 ).
(ii) The RO plant capacity, 𝐹𝑇 (m3 /h).
(iii) The total rated power consumption of the process, 𝑃𝑇
(kW).
(iv) The volume of the seawater tank, 𝑉𝑆 (m3 ).
(v) The battery capacity, 𝐶𝑃 (Ah).
For correct sizing, we assume that a significant record of
data measured in the location is available [11]. Sizing of the
proposed facility can be carried out based on the rated power
PW of the WECs and the measured data.
(i) Using the measured data, the evolution of the power
produced by the WECs 𝑃𝑊(𝑡) is estimated.
(ii) The total rated power consumption of the process 𝑃𝑇
is selected to ensure that one or more of the different
sections of the RO operate for at least 𝜆% of the time.
5
To C + B
DOD ≥ 0.5
L ≥ L max
DOD ≥ 0.4
DOD ≥ 0.35
Stage A
(iii) The tank size 𝑉𝑆 is selected to ensure autonomy at all
times.
(iv) The battery capacity 𝐶𝑃 is selected to ensure
𝑊𝑎 hours of autonomy at all times and that the DOD
is always between given limits DOD < DOD < DOD.
5. Case Study
As an example, the proposed structure of an RO plant
designed to operate consuming between 95 kW and 665 kW
of electricity is shown in Figure 5; section A is designed to
produce 150 m3 /h with a consumption of 340 kW, whereas
section B is designed to produce 70 m3 /h with a consumption
of 150 kW, and C produces 40 m3 /h and it consumes 95 kW.
The energy consumption of the RO units is about 2.4 kWh/m3
and the plant recovery ratio is of 50%.
There are three seawater pumps (SWPs) that provide
water from the seawater intake to the seawater tank. Each
pump works with a power of 45 kW and a flow rate of 80 m3 /h.
These pumps are on-off pumps that work (or not) while the
tank level is between certain values; if the tank level is higher
than a safety value, all the pumps immediately switch off.
Thus, the tank level can be controlled as has been detailed in
Section 3.
6. Results and Discussion
To validate the proposed offshore platform and the sizing
methodology, a case study was carried out for a specific
6 Advances in Mechanical Engineering

×103

Power produced (kW)

Fresh water (m3 /month)

120 90 1500
118 80

On-off SWP/RO
116 70 1000
114 60
112 50
110 500
108 40
106 30 0
104 20 0 20 40 60 80 100 120 140 160
102 10
100 0 Time (days)
0 500 1000 1500 2000 2500 3000 3500
Battery capacity (Ah) Figure 11: Power produced by the WEC.

Water produced (m3 /month)

On-off SWP

Flow rate (m3 /h)

150
On-off RO
100
Figure 9: Effect of battery capacity on system performance (𝑉𝑆 =
50
5500 m3 ).
0
0 20 40 60 80 100 120 140 160
Time (days)
×103
120 30 Section A
Fresh water (m3 /month)

118
116 25 Section B
On-off SWP/RO

114 20 Section C
112
110 15 Figure 12: Fresh water produced in each RO section.
108
106 10
104 5
102 Table 1: Selected operating point.
100 0
0 2000 4000 6000 8000 10000
Tank volume (m3 )
Battery Tank
Fresh water On-off SWP On-off RO
capacity volume
(m3 /month) (times/month) (times/month)
Water produced (m3 /month)
3
(Ah) (m )
On-off SWP 2400 7700 106693 4.8 20.6
On-off RO

Figure 10: Effect of seawater tank volume on system performance

(𝐶𝑝 = 2400 Ah).
location in the Atlantic Ocean, selected for its good wave
regime. For this location, buoy data was measured for 165 days
(wave heights, periods, water salinity, and temperature, etc.).
Based on the available data, a platform was designed using
the sizing methodology proposed in Section 4. This design
was then tested assuming different platform parameters.
The simulation was modeled and performed using MAT-
LAB and Simulink.
(i) Figure 9 shows the effect of battery capacity on system
performance: RO plant capacity and switch on-off
times of the SWPs and the different sections of the
RO for a tank volume (𝑉𝑆 ) of 5500 m3 . It shows that
the amount of fresh water produced increases until a
battery capacity value of 2400 Ah. It remains constant
after this value. The value of the number of times
that the SWPs are switched on-off is independent of
the battery capacity. The number of times that the
RO is switched on-off decreases with the increase of
the battery capacity. Therefore, a battery capacity of
2400 Ah will be chosen to evaluate the effect of tank
size on system performance.
(ii) Figure 10 shows the effect of the seawater tank volume
with the same system parameters. It shows that the
amount of fresh water produced is independent of the
seawater tank volume. The same behavior has been
seen with the number of times that the sections of
the RO are switched on-off. As expected, the number
of times that the SWPs are switched on-off decreases
with the battery capacity.
Finally, an operating point was selected, which is pre-
sented in Table 1.
Figures 11 to 16 show different parameters at the operating
point on 165 days with buoy measurements. It can be seen that
the proposed system operates adequately.
(i) Figure 11 represents the power produced by the WEC.
Power produced depends on the values of the wave’s
height and period (see Figure 3).
(ii) Figure 12 shows that the value of the fresh water pro-
duced in each RO section varies along time according
to the stages of the RO illustrated in Figure 7.
(iii) Figure 13 shows the total fresh water produced, which
is the sum of sections A, B, and C. It can be seen that
this value increases if the value of power produced
(Figure 11) is high.
Advances in Mechanical Engineering 7

300 proposed facility which consists of a desalination system, a

Flow rate (m3 /h)

250 seawater supply, and short-term energy storage, a controller

200
150 has been proposed, and some guidelines for sizing the plant
100 for a specific location have been given. Using buoy data at
50 the proposed location and a multicomponent model-based
0
0 20 40 60 80 100 120 140 160 simulation makes it possible to predict the water production.
Time (days) This proposal has been evaluated for one specific location
(Atlantic Ocean), using buoy-measured data for sizing and
Figure 13: Total fresh water produced by the RO sections. predicting the production at this location. The design was
then evaluated showing how the proposed design method-
ology makes it possible to desalinate water in a sustainable
Power consumed (kW)

800 way. Further work must be done to optimize the sizing of

600 components based on an advanced control system.
400
200 Acknowledgments
0
0 20 40 60 80 100 120 140 160 This work was partly supported by MiCInn (DPI2010-21589-
Time (days) c05-05) and the European Commission (7th Framework
Figure 14: Total power consumed by the RO sections and SWPs.
Programme, Grant Agreement 288145, Ocean of Tomorrow
Joint Call 2011). The authors also acknowledge GRUPO SETA
S. L. for designing the reverse osmosis system and the Marine
Stored seawater (m3 )

8000 Institute of Ireland for collaboration in buoy measurements.

6000
4000 References
2000
0 [1] P. A. Davies, “Wave-powered desalination: resource assessment
0 20 40 60 80 100 120 140 160 and review of technology,” Desalination, vol. 186, no. 1–3, pp. 97–
Time (days) 109, 2005.
[2] D. C. Hicks, G. R. Mitcheson, C. M. Pleass, and J. F. Sale-
Figure 15: Stored seawater in the seawater tank.
van, “Delbouy: ocean wave-powered seawater reverse osmosis
desalination systems,” Desalination C, vol. 73, pp. 81–94, 1989.
Battery charge (%)

100
[3] A. J. Crerar, R. E. Low, and C. L. Pritchard, “Wave powered
80
desalination,” Desalination C, vol. 67, pp. 127–137, 1987.
60
40 [4] A. Serna, D. Torrijos, F. Tadeo, and K. Touati, “Evaluation of
20 wave energy for a near-the-coast offshore desalination plant,”
0 in Proceedings of the World Congress on Desalination and Water
0 20 40 60 80 100 120 140 160 Reuse (IDA), Tianjin, China, October 2013.
Time (days) [5] M. Folley and T. Whittaker, “The cost of water from an autono-
Figure 16: % Battery charge (100%-DOD in percentage). mous wave-powered desalination plant,” Renewable Energy, vol.
34, no. 1, pp. 75–81, 2009.
[6] M. Folley, B. Peñate Suarez, and T. Whittaker, “An autonomous
wave-powered desalination system,” Desalination, vol. 220, no.
(iv) Figure 14 depicts the total power consumed. It is 1–3, pp. 412–421, 2008.
proportional to the total fresh water produced plus [7] F. Tadeo, R. Val, L. Palacin, C. de Prada, and J. Salazar,
the power consumed by the SWPs. “Control of reverse osmosis plants using renewable energies,”
(v) Figure 15 shows the stored seawater. The value is in Proceedings of the 11th IASTED International Conference on
between the minimum and maximum safety levels. Control and Applications (CA ’09), pp. 9–14, Cambridge, UK,
July 2009.
(vi) Finally, Figure 16 depicts the % battery charge. As
[8] U. Seibert, G. Vogt, C. Brennig, R. Gebhard, and F. Holz, “Auton-
expected, this value varies between 30% and 100%; omous desalination system concepts for seawater and brackish
therefore, it never gets discharged. water in rural areas with renewable energies—potentials, tech-
nologies, field experience, socio-technical and socio-economic
7. Conclusions impacts—ADIRA,” Desalination, vol. 168, no. 1–3, pp. 29–37,
2004.
Using wave energy to produce desalinated water in offshore [9] C. Charcosset, “A review of membrane processes and renewable
plants is studied, based on a system completely isolated energies for desalination,” Desalination, vol. 245, no. 1–3, pp.
from the electrical grid. Power consumption adapts to power 214–231, 2009.
production by connecting or disconnecting sections of the [10] L. G. Palacin, F. Tadeo, C. de Prada, and S. Johanna, “Operation
reverse osmosis and the seawater pumps. After presenting the of desalination plants using renewable energies and hybrid
8 Advances in Mechanical Engineering

control,” Desalination and Water Treatment, vol. 25, no. 1–3, pp.
119–126, 2011.
[11] A. Serna and F. Tadeo, “Offshore hydrogen production from
wave energy,” International Journal of Hydrogen Energy, 2013.
[12] E. D. Stoutenburg, N. Jenkins, and M. Z. Jacobson, “Power
output variations of co-located offshore wind turbines and wave
energy converters in California,” Renewable Energy, vol. 35, no.
12, pp. 2781–2791, 2010.
[13] A. Babarit, J. Hals, M. J. Muliawan, A. Kurniawan, T. Moan, and
J. Krokstad, “Numerical benchmarking study of a selection of
wave energy converters,” Renewable Energy, vol. 41, pp. 44–63,
2012.
[14] R. Henderson, “Design, simulation, and testing of a novel
hydraulic power take-off system for the Pelamis wave energy
converter,” Renewable Energy, vol. 31, no. 2, pp. 271–283, 2006.
[15] M. Eriksson, Modelling and experimental verification of direct
drive wave energy conversion: buoy-generator dynamics [Ph.D.
thesis], Uppsala Universitet, Uppsala, Sweden, 2007.
[16] M. Ruellan, H. Benahmed, B. Multon, C. Josset, A. Babarit, and
A. Clement, “Design methodology for a SEAREV wave energy
converter,” IEEE Transactions on Energy Conversion, vol. 25, no.
3, pp. 760–767, 2010.
[17] T. W. Thorpe, “A brief review of wave energy,” Tech. Rep. ETSU-
R120, UK Department of Trade and Industry, Harwell, UK,
1999.
[18] L. G. Rodrı́guez, “Renewable energy applications in desalina-
tion: state of the art,” Solar Energy, vol. 75, no. 5, pp. 381–393,
2003.
[19] M. Wilf, The Guidebook to Membrane Desalination Technology,
Balaban Desalination Publication, L’Aquila, Italy, 2007.
[20] N. Sharmila, P. Jalihal, A. K. Swamy, and M. Ravindran, “Wave
powered desalination system,” Energy, vol. 29, no. 11, pp. 1659–
1672, 2004.
[21] O. Antonia and G. Saur, “Wind to hydrogen in California: case
study,” Tech. Rep. NREL/TP, 5600-53045, National Renewable
Energy Laboratory, 2012.
[22] J. Szymborski, “Lead-acid batteries for use in submarine appli-
cations,” in Proceedings of the 2002 Workshop on Autonomous
Underwater Vehicles, pp. 11–17, June 2002.
[23] I. Yahyaoui, S. Sallem, M. A. B. Kamoun, and F. Tadeo,
“Fuzzy energy management of an off-grid PV/battery system,”
in Proceedings of the International Renewable Energy Congress
(IREC ’12), Hammamet, Tunisia, December 2012.
[24] S. Sallem, M. Chaabene, and M. B. A. Kamoun, “Energy man-
agement algorithm for an optimum control of a photovoltaic
water pumping system,” Applied Energy, vol. 86, no. 12, pp. 2671–
2680, 2009.